Source
- Frijda, Ridderinkhof & Rietveld (2014), LIT-540, read in NOTE-433: “Multiple emotions can exert control”, “Emotion regulation and impulse control”, and the Conclusion’s third and fourth points (C5).
- Barrett (2017), LIT-542, read in NOTE-434: the list of claims not made, Figure 4C, and the Conclusions (C8).
The question
Gross’s founding review (LIT-554, NOTE-429) defines regulation as “the processes by which individuals influence which emotions they have, when they have them, and how they experience and express these emotions” (p. 275), and sorts it by the stage of generation it acts on. His first open challenge is whether regulation can be told apart from generation at all. “Some theorists argue that emotion regulation is best seen as part and parcel of emotion (Frijda, 1986).” He keeps the distinction, and asks for “high” thresholds before regulation is inferred, because inferring it needs an unregulated course to compare against (p. 286; C6 in NOTE-429).
What was actually shown
Frijda’s argument. Events draw several appraisals at once, and each appraisal elicits its own readiness. Appraising what acting on an impulse would lead to is one such appraisal, and the readiness it elicits can check the first: “Anger is often held back by fear of retaliation or loss of affection”. Hence: “There is no clear distinction between processes of emotion and processes of emotion regulation. In large measure, emotion regulation also proceeds impulsively: effortless, by the interaction of concurrent impulsive, emotional processes.” Emotions “are moderated or controlled to the extent that one emotionally cares about the affective consequences of their being uncontrolled”. The paper presents this as “an alternative” to dual-process models (Strack & Deutsch). They “should not, in general, be seen as the outcomes of reason controlling emotion, or the rider controlling the horse … our perspective accords with Hume (1739), in viewing reason as the slave of the passions”. Reflection is not removed. It “seek[s] to help solving impulsive dilemmas that impulsive processes alone are unable to solve”. The support is everyday examples, plus a diary finding that 31% of reported emotions were multiple (Oatley & Duncan 1992, cited). Which state wins depends “upon their relative strengths”, which the paper does not specify further.
Barrett’s convergence. Barrett says this from inside the rival theory of emotion (THEORY-057 against THEORY-055). “Automaticity and control are different brain modes (each of which can be achieved with a variety of network configurations), not two battling brain systems.” She offers as a hypothesis (C8) that “reappraisal and other regulation processes … are accomplished with predictions that categorize sensory inputs and control action with concepts”, which is the mechanism she gives for constructing an emotion. Her Figure 4C shows overlap between the regions active in emotional experience and in regulation (anterior insula, vlPFC, MCC, SMA, posterior STS), with reappraisal drawing more on the frontoparietal network. She reads that difference as the modification of predictions, not as a second system.
So two accounts that disagree about what an emotion is agree that regulation is not a different kind of process. Frijda supports the whole claim. Barrett supports the clause about there being no distinct regulating system, and does not say that regulation is one emotion checking another.
The philosophy-of-mind holdings
- Minsky (LIT-432, NOTE-348) reaches the same architecture by design. Emotional states and Ways to Think are one kind of thing, a selection of resources. One emotion can be used to switch off another, as when anger is used against sleep (§3-8, as NOTE-348 reports it). Control is done by Critics, among them Suppressors and Self-Controllers, which belong to the same machinery. “Free will” is “a process we use to stop other processes”. This is a design proposal argued by example, so it is a neighbour and not a source.
- Brooks (LIT-436, NOTE-351) builds it. A newer behaviour layer steers an older one only by overriding its signals for a while, and the robot “is a collection of competing behaviors”. The layers are ordered, but the overriding layer is another behaviour and not a reasoner. That is control without a controller of a different kind, which is what this account claims for emotion.
- Haidt (LIT-511, NOTE-403) has the same shape for moral judgement. Reasoning rarely overrides intuition (link 5). Its private influence runs mostly through role-taking, which triggers a competing intuition (link 6). Frijda’s reflection likewise works by raising a competing readiness. Both cite Hume.
- Self-determination theory argues the contrary (LIT-558, NOTE-439). Deci and Ryan (2000) hold that needs, not emotions, are the higher-order regulators of behaviour, “because emotions themselves must be self-regulated for effective functioning, and the basic psychological needs are centrally involved in the processes by which this self-regulation occurs” (p. 254). That posits a regulating level above emotion, which this account denies. The claim is argued, not tested, and no record THEORY states it, so it is not declared a rival. It may be closer than it looks: Frijda’s concerns, which give events their pertinence, include the person’s needs (NOTE-433).
- THEORY-029 (higher-order attitudes and ownership). Hierarchical accounts of agency place the agent’s authority in an attitude of a higher order than the motive it governs. If this account is right, there is no such level in the psychology of control. A motive prevails by its relative strength and by the concerns it recruits, and all of these are first-order. This does not decide THEORY-029’s question, which is about whose a motive is and not about which one wins. It does mean that a hierarchical account cannot borrow its structure from how self-control works. Frijda’s lever, recruiting “non-obvious but relevant concerns, such as moral concerns and self-esteem”, puts what one cares about in the place Watson’s values occupy, and THEORY-029 records that Velleman’s dilemma reaches values too. The bearing is the record’s. Neither Frijda nor Barrett discusses autonomy.
- THEORY-040 (manipulation and history) is not touched. This account is about present structure only.
Three accounts filed since. Two support the clause that there is no
regulator above the competing states, from outside emotion research.
Ainslie’s bargaining (THEORY-067) makes self-control the interaction of
interests that dominate at different delays, with no faculty of will over
them; it also offers an answer to which state wins, though what does the
checking there is an expectation about one’s own future choices rather than
a concurrent emotion. Arbitration by reliability (THEORY-062) lets
model-based and model-free controllers compete as peers, the more reliable
estimate winning, not the one a higher system prefers. The expected value of
control (THEORY-070) is the challenge, and a conditional one: it posits a
dedicated system that decides how much control to apply, which would be the
distinct system this account says refutes it only if that specification
were shown to govern emotion regulation generally, effortless regulation
included. The test is the one promote_when already names, whether
effortless regulation by a competing emotion works without recruiting the
control networks that deliberate regulation recruits.
What this does not say
- It does not say that regulation never happens, or that deliberate strategies do nothing. Gross’s cited contrast, in which reappraisal lowers negative experience and suppression does not reliably do so and raises sympathetic activation (C3 in NOTE-429), stands. Strategies differ in their consequences whether or not a separate system runs them.
- “Most of it” is the claim. Frijda says “in large measure”, and he keeps a role for reflection. A rare case of reflection overriding an impulse directly would not refute the account. A distinct system that does the work generally would.
- It does not say which state wins. Neither source predicts that, and it is the account’s largest gap.
Note of 2026-10-28: what holds it back
Kept Proposed. Both sources argue it and neither tests it (NOTE-433 C5,
NOTE-434 C8). The record holds no result that promote_when names: no
dissociation of regulation from generation, no measured regulation without
a competing concern, and no evidence that reappraisal works through a
second, competing readiness or that effortless regulation spares the
control networks. THEORY-878’s regions selective for reappraisal are a
difference in network activation, which promote_when says cannot settle
it, and THEORY-871 fits either account. Gross and Barrett (2011) remains
unfiled.
Note of 2026-10-30: what holds it back
No result that promote_when names is held. There is no dissociation of
regulation from generation, no measured regulation without a competing
concern, and no evidence that reappraisal works through a competing
readiness or that effortless regulation spares the control networks. Gross
and Barrett (2011) is still unfiled. THEORY-871 was revised on 2026-10-30
only in its edge to CLAIM-412, which became corroborates (ADR-058), and its
result still fits either account.